Convenient New Chemical Actinometer Based on Aqueous Acetone, 2

Tucson, Arizona 85721. A convenient new chemical actinometer was developed to measure the spectral output of laboratory ultraviolet (UV) light sources...
0 downloads 0 Views 125KB Size
Environ. Sci. Technol. 2005, 39, 2262-2266

Convenient New Chemical Actinometer Based on Aqueous Acetone, 2-Propanol, and Carbon Tetrachloride H E N G L I , † E R I C A . B E T T E R T O N , * ,‡ ROBERT G. ARNOLD,§ WENDELL P. ELA,§ BRIAN BARBARIS,‡ AND CECILIO GRACHANE§ Department of Civil and Environmental Engineering, University of Michigan, 2200 Bonisteel Boulevard, Ann Arbor, Michigan 48109, Department of Atmospheric Sciences, The University of Arizona, P.O. Box 210081, Tucson, Arizona 85721-0081, and Department of Chemical and Environmental Engineering, The University of Arizona, Tucson, Arizona 85721

A convenient new chemical actinometer was developed to measure the spectral output of laboratory ultraviolet (UV) light sources over the wavelength range of 260-330 nm. It can also be used to measure solar UV irradiance (e325 nm). The actinometer is based on the photoreduction of aqueous carbon tetrachloride (CT) to chloroform (CF) in the presence of acetone (the chromophore) and 2-propanol (the reductant). In all cases, CT disappearance (and CF formation) followed zero-order kinetics over 95% of the reaction. The slope of the linear decay curve forms the basis of the new actinometer, which was calibrated using ferrioxalate actinometry. Quantum yields were measured at 10 nm intervals and were found to be uniform throughout the range of 260-300 nm. As expected, quantum yields gradually decreased to zero as the wavelength was increased from 300 to 340 nm. The high quantum yields (≈150), low sensitivity to room light, and the straightforward determination of [CT] and [CF] by gas chromatography offer significant advantages over some other chemical actinometers, which might require the preparation and purification of lightsensitive compounds in a darkened environment and long exposure times.

Introduction The determination of chemical quantum yield is fundamental to the study of a photochemical reaction. Quantum yield is a measure of the efficiency of the photoreaction and can reveal the existence of chain reactions and other mechanistic details. When it comes to environmental phototreatment systems, high quantum yields are desirable in order to achieve high treatment efficiency and keeping energy consumption low, especially when photons are generated electrically. * Corresponding author phone: (520)621-2050; fax: (520)621-6833; e-mail: [email protected]. † Department of Civil and Environmental Engineering, University of Michigan. ‡ Department of Atmospheric Sciences, The University of Arizona. § Department of Chemical and Environmental Engineering, The University of Arizona. 2262

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

Quantum yield is the ratio of chemical reaction rate to absorbed photon flux and thus requires the simultaneous measurement of two quantities. The first (chemical reaction rate) is often easy to measure, but the second (photon flux) is often more difficult. In many laboratories, chemical actinometry is used to determine photon flux. The advantages over instrumental methods include less complicated calibration, less time-consuming and less expensive procedures, and the ability to match the geometry of the actinometer solution with that of the light source. Currently, the most widely accepted liquid-phase chemical actinometer employs aqueous potassium ferrioxalate (1-4). This actinometer has been extensively studied (e.g., see refs 5-12 and references therein). Briefly, upon illumination a ligand-to-metal charge transfer occurs, resulting in the oxidation of the oxalate and reduction of Fe(III) to Fe(II), which is then determined colorimetrically using o-phenanthroline. Quantum yields are close to unity over a wide wavelength range. The advantages of this actinometer lie primarily in its sensitivity and reliability. The method also benefits from a linear dynamic wavelength range from 253.7 to 577 nm, over which the system can be utilized. Nevertheless, because ferrioxalate absorbs visible light, operations including synthesis of the ferrioxalate and colorimetric determination of Fe(II) must be carried out in a darkroom. Furthermore, errors can arise from slow color development of the Fe(II)-phenanthroline complex due to photooxidation of the phenanthroline solution and competitive complexation of the phenanthroline by Fe(III) (13). A second liquid-phase chemical actinometer is based on the photoisomerization of o-nitrobenzaldehyde to o-nitrobenzoic acid over the range of 310-400 nm and subsequent measurement of time-dependent hydrogen ion or nitrobenzaldehyde concentration (14, 15). This method is unique in that it offers an experimental protocol in which photon flux can be measured with a pH meter. However, when used in this manner, actinometer accuracy is limited by the sensitivity of pH measurements and potential interference from the absorption of atmospheric CO2 when the pH is close to neutral. To remedy these and other limitations, we tested a new actinometer based on the photoreduction of carbon tetrachloride (CT) to chloroform (CF) in aqueous acetone (AC) in the presence of 2-propanol (iP). We have found it to be useful in the near-UV range (260-330 nm) where many solar-driven photochemical reactions are important. We evaluated the new system based on the following criteria: (i) the chemicals involved are easy to acquire and safe to handle in room light; (ii) the samples can be easily analyzed by commonly available methods; (iii) only one species absorbs incident light, and the wavelength of the light falls in the near-UV range; and (iv) the quantum yields are high, constant, and insensitive to small variations in room temperature, solution pH, and actinometer concentrations. CT is photochemically reduced to CF via the following reactions when a solution of CT in aqueous acetone and 2-propanol is irradiated in the near-UV (16): hν

(CH3)2CO9 8 (CH3)2CO* φ

(1)

(CH3)2CO* + (CH3)2CHOH f 2(CH3)2CHO•

(2)

(CH3)2CHO• + CCl4 f (CH3)2CO + CHCl3 + HCl (3) 10.1021/es050046y CCC: $30.25

 2005 American Chemical Society Published on Web 02/24/2005

Measurement of -d[CT]/dt and/or d[CF]/dt offers the potential for a convenient actinometer if the spectral quantum yield can be determined independently. Here we report the calibration of the spectral output of a Xe arc lamp by ferrioxalate actinometry, which in turn allowed us to measure the desired spectral quantum yield for the proposed AC/CT/2-propanol actinometer.

Experimental Section Chemicals and Materials. All chemicals were from commercial sources and used without further purification (acetone: Fisher, HPLC grade; 2-propanol: Mallinckrodt, AR; carbon tetrachloride: ACROS, 99.8%). Solutions were prepared with distilled and deionized (DI) water obtained from a Millipore Milli-Q water purification system (resistivity >18 MΩ‚cm). Aqueous 2-propanol/acetone solutions were prepared volumetrically to produce the desired concentrations and stored in the dark. They were used within 1 week. All glassware was routinely water-washed, soaked overnight in 2% cleaning solution (MICRO-90), flushed and rinsed with DI water, and baked overnight at 500 °C prior to use. Potassium ferrioxalate was prepared from K2C2O4 (Fisher, ACS grade, 99.7%) and FeCl3 (Fisher, laboratory grade) (4). The reagent was stored in a dark bottle tightly wrapped with aluminum foil, and the bottle was stored in the dark. Other reagents (FeSO4‚7H2O: Fisher, ACS crystal; H2SO4: Fisher, ACS; NaC2H3O2‚3H2O: Ashland, ACS; 1,10-phenanthroline: LabChem, 0.1% w/w APHA) were used as received. Colorimetric analysis showed that the ferrioxalate contained traces of Fe(II) (0.1%, molar ratio). This amount of Fe(II) did not significantly interfere with the absorbance readings. The blank (un-irradiated solution) was used in the reference beam. The quantum yield of Fe(III) conversion was also unaffected by the trace Fe(II) impurity because only about 1% of the initial Fe(III) was reduced to Fe(II) during irradiation. Light Source. The light source was a 1000 W xenon arc lamp mounted in a lamp housing with an f/4 ellipsoidal reflector (Photon Technology International, A-6000). The power supply was a regulated DC unit (PTI model LPS 1200) that provided stable current ( 2; where [AC] is the acetone concentration (M), λ (M-1 cm-1) is the molar absorptivity of acetone at the selected wavelength, λ; and b is the path length (cm). The collimator lens eliminated reflective/refractive losses at the cylinder walls and a black cloth under the bottom of the cylinder further reduced any possible reflection. Small liquid samples (25 µL) were withdrawn from the side port with a gastight syringe and injected into standard 2-mL amber GC vials containing 1.00 mL 2-propanol as solvent. Triplicate sampling was performed at each designated time point. CT and its photoreduction product (CF) were determined by gas chromatography (Hewlett-Packard 5890A; 7673 auto injector and controller; 3392A integrator; HP-5 cross-linked 5% Ph Me silicone column 30 m × 0.53 mm × 2.65 µm; isothermal 60 °C; 120 °C injector and detector; electron capture detector; He carrier, column head pressure 50 psi; injection volume 1.0 µL). Under these conditions, the retention times for CT and CF were 2.4 and 1.8 min, respectively. The instrument was calibrated with gravimetrically prepared standards (F ) 0.785 g mL-1). Linear calibration curves were obtained over the experimental range (r2 ) 0.9985). Measured CT concentrations were plotted versus time of irradiation, and the photoreaction rate constants were derived from the slope. Results of a typical experiment are shown in Figure 1. There is an initial lag phase when residual O2 consumes electrons that would otherwise be available for CT reduction. The zero-order rate constant, kλ (mol L-1 min-1), was calculated from the linear regression slope of the linear portion of the curve (i.e., excluding the lag phase). The rate constant was normalized by dividing by the initial CT concentration [CT]0 (mol L-1) to give the specific zeroorder rate constant, kλ0 (min-1). Equivalently, kλ0 can be VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2263

obtained by plotting the relative concentration [CT]λ/[CT]0 versus time (in all cases, r2 > 0.99):

kλ ) d[CT]λ/dt 0

kλ ) kλ/[CT]0 ) d([CT]λ/[CT]0)/dt The actinometer was operated at room temperature, which ranged from 22 to 28 °C. The quantum yield in a light-limited system such as ours is unaffected by such a small temperature variation (21, 23, 24). Actinometry with Potassium Ferrioxalate. The spectral output of the Xe lamp was measured using the potassium ferrioxalate actinometer (1-4) as our primary standard. All quantitative work relating to the handling of the ferrioxalate solutions was carried out in a darkroom, using a red photographic safe light. Irradiation time was adjusted to produce approximately 0.06 mM Fe(II) from the original 6 mM K3Fe(C2O4)3 solution (1% conversion). This amount of Fe(II), when fully complexed with phenanthroline, resulted in absorbance readings at 510 nm in the range of 0.5-1.0 on a dual beam spectrophotometer (1-cm quartz cuvette, using the un-irradiated solution as reference; Shimadzu, UV2101PC). A sufficiently large amount of phenanthroline was added ([phenanthroline]/[Fe(II) > 50:1) to overcome potential errors from slow color development. We determined the molar absorptivity (base 10) of the ferrous-phenanthroline complex to be 1.13 × 104 M-1 cm-1 at 510 nm, which was used in our calculation. This is close to the literature value of 1.11 × 104 M-1 cm-1 (4). The photon flux, in 10-nm intervals from 260 to 340 nm, was determined based on a quantum yield (Φλ) of 1.24 for Fe(II) formation (4). The following equation, which is from ref 4, was used to calculate the absorbed spectral flux:

Iabs )

V2 × A [Fe2+] ) einstein L-1 min-1 t × Φλ t × Φλ × V1 × l × 510

where [Fe2+] is the amount of Fe2+ produced (mol/L) during the irradiation time period (t); V1 is the aliquot volume taken from the irradiated solution for reaction with phenanthroline; V2 is the volume of Fe2+-phenanthroline complex development solution used for spectrophotometric measurement at 510 nm; A is the absorbance reading at 510 nm; l is the path length of the cuvette (1 cm); and 510 is molar absorptivity at 510 nm (1.13 × 104 M-1 cm-1). A plot of Fe(II) versus time gave Iλ (mol L-1 min-1), the absorbed spectral flux. All of the incident light was absorbed by the ferrioxalate solutions (Aλ > 2), therefore I0 ) Iabs. The data are summarized in Table 1. Since both the ferrioxalate and AC/CT/iP systems were optically opaque and the same volume of solution and the same geometry was used for both systems, any reflective/refractive losses at the solution surface or during passage through the solution would be identical. Consequently, Iλ is identical for both systems. Also, assuming that losses at the quartz window and at the air/solution interface are small, Iλ ≈ I0λ, the spectral output of the Xe lamp system.

Results and Discussion In all experiments, relatively high initial CT concentrations (2.0-3.0 mM) were chosen to maintain excess CT over a significant period of reaction. CT reduction followed zeroorder kinetics until [CT] fell below 0.1 mM (i.e., below our CT detection limit under these conditions). The results of a typical experiment are shown in Figure 1. After an initial lag phase of about 50 min there is a zero-order (linear) decay of CT and stoichiometric production of CF, as expected (16). The initial lag is due to the photoreduction of O2 (eq 4), 2264

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

TABLE 1. Summary of the Quantum Yields of CT Photoreduction to CF in the Aqueous Acetone-CT-2-Propanol Actinometera wavelength (nm)

kλ0 (min-1)

kλ (mM min-1)

Iλb (einstein L-1 min-1)

Φλc

235-245 245-255 255-265 265-275 275-285 285-295 295-305 305-315 315-325 325-335 335-345

0 1.15 × 10-3 1.41 × 10-3 1.96 × 10-3 2.42 × 10-3 3.73 × 10-3 5.10 × 10-3 5.56 × 10-3 4.74 × 10-3 2.57 × 10-3 0

0 0.345 × 10-2 0.423 × 10-2 0.588 × 10-2 0.726 × 10-2 1.119 × 10-2 1.560 × 10-2 1.668 × 10-2 1.422 × 10-2 0.771 × 10-2 0

2.619 × 10-8 3.700 × 10-8 4.754 × 10-8 7.504 × 10-8 9.605 × 10-8 11.706 × 10-8 13.292 × 10-8 14.878 × 10-8 18.092 × 10-8

162 ( 14 159 ( 7 153 ( 5 149 ( 7 162 ( 6 142 ( 13 107 ( 2 52 ( 1 0

a [CT] 0 ) 3.0 mM; actinometer solution volume ) 236.3 mL. Obtained from ferrioxalate actinometry; Iλ is not known for λ < 255 nm. c Calculated from eq 5. Uncertainties were estimated by the error propagation rule for division operation. b

FIGURE 2. Electronic spectra (base 10) of species used in the chemical actinometer. Pure water and 2-propanol do not absorb light at wavelengths above 200 nm (data not shown). Acetone, CT, and CF were diluted with aqueous 5.7 M 2-propanol to produce absorbance readings in the range of 0.5-1.0. The overlap of the acetone absorbance with the reported solar spectrum at the earth’s surface makes this new chemical actinometer useful for measuring solar irradiance in the ultraviolet below 325 nm. As indicated by the arrow, wavelength-dependent solar fluxes (solid line) are on the right-hand vertical axis. All other curves represent wavelengthdependent molar absorptivities (left-hand vertical axis). which has been studied in detail (17, 18). The duration of the lag phase depends on the oxygen concentrationsonce all the oxygen is consumed, CT reduction begins. Therefore, the initial oxygen content does not affect the subsequent CT reduction: hν

2(CH3)2CHOH + O2 98 2(CH3)2CHO + 2H2O

(4)

Rates of CT disappearance and CF formation were identical within experimental error, indicating that no significant secondary reactions (e.g., CF f dichloromethane) proceeded under these conditions. Also, the good carbon balance (CT + CF) indicates that there were no significant byproducts. The absorption spectra of all chemical species present in the actinometer are shown in Figure 2. The onset wavelength for absorption for acetone is about 320 nm. Over the range 320 nm to 250 nm, only acetone exhibited strong absorptivity with a maximum at 267 nm (267 ) 17.5 M-1 cm-1) due to the n f π* transition. Water and 2-propanol are transparent over this entire range ( < 0.001 M-1 cm-1, data not shown). Consequently, even at high concentrations, 2-propanol does not absorb a significant fraction of the incident light. Both CT and CF begin to absorb at shorter wavelengths (e.g., at

FIGURE 3. CT disappearance rate profiles for the AC/iP aqueous reaction system as a function of irradiation wavelength. Each wavelength listed corresponds to the center of the 10 nm bandpass window. (Initial [iP] ) 5.7 M; initial [CT] ) 2.2 mM; AC concentration was adjusted to ensure more than 99% light absorption; optical path length ) 22 cm; purged with argon for 30 min.)

actinometer to measure Iλ for other light sources simply by rearranging eq 5. The uncertainty of Φλ values reported in Table 1 was estimated from percent relative uncertainties of both the CT loss rates and the photon flux measurements using the division error propagation rule. Relative uncertainties in CT loss rates and photon fluxes were calculated based on results from repeated experiments at individual wavelength intervals. Values of the quantum yields were remarkably constant in the range of 260-310 nm. It seems likely that loss of reactivity (Figure 3) as the wavelength falls below 240 nm is due to decline in lamp output. On the other hand, low acetone absorptivity was responsible for the declining efficiency as wavelengths exceed 310 nm (Figure 2). The high quantum yields of CT photoreduction are attributed to the involvement of free radical chain reactions. Triplet acetone excited from the direct photon absorption abstracts a hydrogen atom from a 2-propanol molecule, generating two identical 1-hydroxyl-1-methylethyl-ketyl radicalssone from the alcohol and one from the ketone. This radical is strongly reducing (22). Reduction of CT to CF is accompanied by Cl- elimination and H+ production. Although variation in pH from ∼6.5 to 3.5 was observed over the course of an irradiation experiment (data not shown), CT degradation rates were unaffected. This confirms that the pH change incurred by the photoreduction does not interfere with the quantum yield or rate of the light-limited reaction. This aqueous system of acetone-2-propanol-CT, owing to its simplicity, flexibility, and reliability, should be very useful for determining the light intensities of many photochemical investigations.

Acknowledgments

FIGURE 4. Spectral quantum yields (Φλ) of CT degradation in the acetone-2-propanol-CT aqueous system. The solid curve is an arbitrary fit. 250 nm and [CT] ) 3.0 mM, about 0.8% incident light is absorbed by CT). Figure 2 also shows the overlap of the AC/ CT/2-propanol actinometer with the near-UV of the solar spectrum. The effect of varying wavelengths over the range of 240340 nm at 10-nm intervals on the values of kCT0 is illustrated in Figure 3. Rate constants are summarized in Table 1. The most important feature is the consistent zero-order kinetics throughout the wavelength range. The highest reaction rates were observed in the range of 290-320 nm, or at somewhat higher wavelengths than the maximum acetone absorptivity. The maximum reaction rate was at 310 nm. Thus, it is evident that rate changes are affected by wavelength-dependent light output from the Xe lamp as well as acetone absorption. The overall spectral quantum yield (Φλ) for CT conversion to CF was calculated from eq 5:

Φλ ) kλ/φλIλ

(5)

Here, Iλ represents the absorbed spectral flux at the 10-nm wavelength interval, centered at λ (einstein L-1 min-1); and φλ is the primary spectral quantum yield for triplet acetone formation, which is unity for the wavelength range used here (19-21). Figure 4 shows the spectral quantum yields for CT degradation calculated using eq 5. These values are also listed in Table 1 for convenience. Where a desired Φλ value is not directly available from Table 1, a simple linear interpolation is recommended. The AC/CT/iPr system can be used as an

This research was made possible by Grant P42 ES004940 from the National Institute of Environmental Health Sciences, NIH, with funding provided by the U.S. EPA. We thank Ms. Laura Berry for her assistance in the laboratories.

Literature Cited (1) Parker, C. A. A new sensitive chemical actinometer. 1. Some trials with potassium ferrioxalate. Proc. R. Soc. London, Ser. A 1953, 220, 104-116. (2) Hatchard, C. G.; Parker, C. A. A new sensitive chemical actinometer. 2. Potassium ferrioxalate as a standard chemical actinometer. Proc. Roy. Soc. London, Ser. A 1956, 235, 518-536. (3) Kuhn, H. J.; Braslavsky, S. E.; Schmidt, R. Chemical actinometry. Pure Appl. Chem. 1989, 61, 187-210. (4) Rabek, J. F. Experimental Methods in Photochemistry and Photophysics, Part 2; John Wiley & Sons: New York, 1982; pp 944-947. (5) Lee, J.; Seliger, H. H. Quantum yield of ferrioxalate actinometer. J. Chem. Phys. 1964, 40, 519. (6) Kurien, K. C. Modification to ferrioxalate actinometer. J. Chem. Soc. B 1971, 2081. (7) Wrighton, M.; Witz, S. Stability of iron(II) in ferrioxalate solutions. Mol. Photochem. 1972, 3, 387-394. (8) Fernandez, E.; Figuera, J. M.; Tobar, A. Use of the potassium ferrioxalate actinometer below 254 nm. J. Photochem. 1979, 11, 69-71. (9) Connolly, J. S.; Meyer, T. H. A convenient irradiation cell for ferrioxalate actinometry. Photochem. Photobiol. 1981, 34, 145146. (10) Demas, J. N.; Bowman, W. D.; Zalewski, E. F.; Velapoldi, R. A. Determination of the quantum yield of the ferrioxalate actinometer with electrically calibrated radiometers. J. Phys. Chem. 1981, 85, 2766-2771. (11) Langford, C. H.; Holubov, C. A. Wavelength and temperaturedependence in the photolysis of the chemical actinometer, potassium tris-oxalatoferrate(III), at longer wavelengths. Inorg. Chim. Acta. Lett. 1981, 53, L59-L60. (12) Nicodem, D. E.; Aquilera, O. M. V. Standardization of the potassium ferrioxalate actinometer over the temperature-range 5-Degrees-C 80-Degrees-C. J. Photochem. 1983, 21, 189-193. VOL. 39, NO. 7, 2005 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

2265

(13) Kirk, A. D.; Namasivayam, C. Errors in ferrioxalate actinometry. Anal. Chem. 1983, 55, 2428-2429. (14) Pitts, J. N., Jr.; Vernon, J. M.; Wan, J. K. A rapid actinometer for photochemical air pollution studies. Int. J. Air Water Pollut. 1965, 9, 595-600. (15) Willett, K. L.; Hites, R. A. Chemical actinometry: Using o-nitrobenzaldehyde to measure light intensity in photochemical experiments. J. Chem. Educ. 2000, 77, 900-902. (16) Betterton, E. A.; Hollan, N.; Arnold, R. G.; Gogosha, S.; McKim, K.; Liu, Z. J. Acetone-photosensitized reduction of carbon tetrachloride by 2-propanol in aqueous solution. Environ. Sci. Technol. 2000, 34, 1229-1233. (17) Coxon, J. M.; Halton, B. Organic Photochemistry, 2nd ed.; Cambridge University Press: London, 1987. (18) Stefan, M. I.; Hoy, A. R.; Bolton, J. R. Kinetics and mechanism of the degradation and mineralization of acetone in dilute aqueous solution sensitized by the UV photolysis of hydrogen peroxide. Environ. Sci. Technol. 1996, 30, 2382-2390. (19) Scaiano, J. C. Intermolecular photoreductions of ketones. J. Photochem. 1973, 2, 81-118.

2266

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 39, NO. 7, 2005

(20) Turro, N. J. Modern Molecular Photochemistry; Benjamin Cummings: Menlo Park, CA, 1978; p 91. (21) Alfassi, Z. B. Chemical Kinetics of Small Organic Radicals, Vol. 3; CRC Press: Boca Raton, FL, 1988; pp 41-69. (22) Wardman, P. Reduction potentials of one-electron couples involving free radicals in aqueous solution. J. Phys. Chem. Ref. Data 1989, 18, 1637-1755. (23) Li, T.; Farrell, J. Electrochemical investigation of the rate-limiting mechanisms for trichloroethylene and carbon tetrachloride reduction at iron surfaces. Environ. Sci. Technol. 2001, 35, 35603565. (24) Scherer, M. M.; Westall, J. C.; ZiomekMoroz, M.; Tratnyek, P. G. Kinetics of carbon tetrachloride reduction at an oxide-free iron electrode. Environ. Sci. Technol. 1997, 31, 2385-2391.

Received for review August 2, 2004. Revised manuscript received January 10, 2005. Accepted January 18, 2005. ES050046Y